A benchmark of in-the-wild distribution shifts spanning diverse data modalities and applications, from tumor identification to wildlife monitoring to poverty mapping.
Learn how to install and use our Python package, which provides a simple and standardized interface for all WILDS datasets.
Read more
WILDS consists of 7 datasets across a diverse range of data modalities, applications, and types of distribution shifts. Explore the datasets here.
Explore datasets
We track the state-of-the-art on each dataset. View and submit your results here.
View leaderboard
WILDS is maintained by researchers from Stanford, Berkeley, Caltech, Cornell, and Microsoft Research. We are actively adding datasets; please contact us if you are interested in contributing.
Read our paper for an overview of distribution shifts in the wild and for more details on each dataset.
Read the paper
WILDS is an open-source community project. Clone the code, report an issue, or ask questions and take part in discussions.
Go to repo